The automotive industry suffered first, in part because it only accounts for 10 per cent of the revenue of chip makers and had less bargaining power than the consumer electronics market, which accounts for about 70 per cent of chip foundries’ revenues, said Michael Schallehn, a partner at management consulting firm Bain & Co, which specialises in the semiconductor industry.
“But the crisis is now expanding across the entire tech industry,” he said. “Automotive was really the canary in the coal mine. The issues that we’ve started to see in automotive we’re expecting to expand to other industries as well, though probably not to the same extent.”
Part of the problem was that the world’s chip foundries were already operating at close to 100 per cent capacity, even before the COVID-19 pandemic caused a 60 per cent spike in demand for consumer electronic devices such as smartphones, laptops and computer screens, said Peter Hanbury, another Bain analyst who specialises in the manufacturing sector.
“You can’t just hire more contract workers and try to increase the output of a factory,” Mr Schallehn warned. “That’s not possible in semiconductor manufacturing ... We expect this shortage to last until 2022, and it’s going to cost billions in order to add new capacity.”
Another problem is simply that chips are hard to make: they take three or four years to design, two to three years to build a foundry capable of making them, and even once all that is done, it still takes two or three months to make a batch of them, in a clean room with highly trained workers, he said.
“There are structural issues inside of semiconductor manufacturing – choke points along the entire supply chain – so this [crisis] could be repeated. Given the industry structure it’s likely that we’ll see shortages again in the near future,” Mr Schallehn said.